检索结果-内蒙古大学图书馆

Leveraging feature fusion ensemble of VGG16 and ResNet-50 for automated potato leaf abnormality detection in precision agriculture

引用

Soft computing 2025年第4期29卷 2263-2277页

作者： Trivedi, Amit Kumar Mahajan, Tripti Maheshwari, Tanmay Mehta, Rajesh Tiwari, Shailendra Department of Computer Science and Engineering Thapar Institute of Engineering and Technology Punjab Patiala147001 India

In the era of advancement in technology and modern agriculture, early disease detection of potato leaves will improve crop yield. Various researchers have focussed on disease due to different types of microbial infection in potato leaves using computer vision and machine learning approaches. In this paper, a data science approach for multiclass classification of potato normal and abnormal leaves due to fungal infection like early blight and late blight is performed using the ensembling of deep learning (DL) CNN models. Firstly, the performance of classification on potato disease is verified separately on VGG16 and ResNet-50 CNN models after pre-processing of the leaf dataset. The pre-processing includes noise removal and normalization. Further improvement in classification accuracy is achieved by the ensembling of VGG16 and ResNet-50 CNN models. The ensembling of CNN models is performed on the feature level by fusing features extracted using VGG16 and ResNet-50. From the experimental results, performed on publicly available datasets consisting of 2152 number of normal and abnormal images it is observed that the average classification accuracy of 98.22%, 96.16% and 95.68% is achieved using the proposed ensemble, VGG16 and ResNet-50 models respectively. The efficacy of the proposed approach (ensemble technique at feature level fusion) is verified in comparison with recently reported DL model-based approaches. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Text steganography: enhanced character-level embedding algorithm using font attribute with increased resilience to statistical attacks

引用

Multimedia Tools and Applications 2025年第11期84卷 9367-9392页

作者： KL, Shreya Narasimhan R, Bala Krishnan Department of Computer Science and Engineering National Institute of Technology Tiruchirappalli Tamil Nadu Tiruchirappalli India

Text steganography, the science of hiding secret messages in innocent-looking text documents ensures the secrecy of the embedded secret. Cryptography, on the other hand, encrypts and converts the secret message into an unintelligible form ensuring confidentiality, integrity, and authentication of the message. It is to be noted that both steganography and cryptography are expected to work together hand-in-hand in order to attain a better security. Hence, a methodology that combines techniques from both steganography and cryptography is an unavoidable one for the complete protection of sensitive data. This paper aims to achieve one such methodology. The best existing method which works in similar direction is observed to use character spacing feature of text documents to embed the secrets. The method is reported to attain an average embedding capacity of ≈2 bits/cover-character and 8-bits/distortion with high imperceptibility and considerable cryptographic security. However, during experimentation, it is noticed that the methodology has the potential vulnerability of leaving clues for attackers with a probability of 0.51% which paves way for possible statistical attacks namely frequency analysis. This vulnerability occurred because of the constant left circular shift operation that is performed after embedding each secret character. This paper aims to address this vulnerability by proposing a novel modification that facilitates the methodology to perform both left and right circular shifts in an unbiased manner, based on the embedded secret, without affecting the notable achievements of the existing methodology. The proposed modification is successfully implemented, and further analysis indicated that the performed modification has in fact embedded the secret characters uniformly thereby reducing the standard deviation by 24%. This increased the resilience of the extended method to potential attacks that rely on statistical techniques. © The Author(s), under exclusive

关键词： Cryptography

来源：评论

学校读者我要写书评

暂无评论

Complex preference analysis: a score-based evaluation strategy for ranking and comparison of the evolutionary algorithms

引用

Soft computing 2025年第4期29卷 1967-1980页

作者： Sarkar, Debojyoti Biswas, Anupam Department of Computer Science and Engineering National Institute of Technology Silchar Assam Silchar788010 India

In this paper, an evaluation strategy is proposed for evaluation of optimization algorithms, called the Complex Preference Analysis, that assesses the efficiency of different evolutionary algorithms by considering multiple performance metrics concurrently across diverse dimensions and generating a single score for ranking. The proposed strategy allows for a thorough evaluation by assigning distinct priorities in the form of weights, to various performance parameters. The algorithm’s performances on each performance parameter are captured and the best and worst performances in each performance category are identified. The deviation of each algorithm from the best and worst performances in the respective performance category is formulated to generate a single score for each of the algorithms for comparison. The flexibility to adjust weights provides a valuable tool for customizing evaluations according to individual needs and preferences. The proposed evaluation strategy is studied by running evaluations on nine evolutionary algorithms based on their performance over five performance indicators on six datasets of the Travelling Salesman Problem. Conducting Complex Preference Analysis on the different evolutionary algorithms reveals distinctive patterns in their performance, with some algorithms such as Simulated Annealing, Tabu Search, Bat Algorithm, Genetic Algorithm, etc. consistently demonstrating superior results across multiple Travelling Salesman Problem instances, while other algorithms such as Ant Colony Optimization, Teaching Learning Based Optimization, etc. rank lower. The algorithms’ performances on individual performance parameters, when reflected on the final rank of the algorithms, generated based on the collective impact of all the parameters, elucidate the efficiency of the proposed evaluation strategy. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Optimization algorithms

来源：评论

学校读者我要写书评

暂无评论

Activation Control of Vision Models for Sustainable AI Systems

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第7期5卷 3470-3481页

作者： Burton-Barr, Jonathan Fernando, Basura Rajan, Deepu Institute of High-Performance Computing The Centre for Frontier AI Research Agency for Science Technology and Research 138632 Singapore Nanyang Technological University Nanyang School of Computer Science and Engineering 639798 Singapore

As artificial intelligence (AI) systems become more complex and widespread, they require significant computational power, increasing energy consumption. Addressing this challenge is essential for ensuring the long-term sustainability of AI technology. AI-on-AI control refers to a system with a set of AI functions controlled by an upper-level AI model. Previous work in AI-on-AI control focuses on boosting accuracy or expanding system capability by increasing overall system cost. Alternatively, we focus on applying AI-on-AI control to decrease system cost and increase the sustainability and viability of a system with multiple AI functions. Our supervised image classification evaluative controller (SICEC) is a cost-reduction oriented AI-on-AI controller that learns when vision models within an AI system should be activated based on input features. The function controller (FC) preprocesses an input and activates relevant functions, functions being distinct units of AI functionality within the system. Some functions have a set of same functional models (SFMs). These models take the same input and produce the same output but have architectural differences. We introduce a same functional controller to select a SFM using the FC's decision confidence. Results are promising, with a decrease of up to 48.9% in inference time, 67.8% in floating point operations (FLOPs), and 66.4% in energy usage. With SICEC showing significant reductions in inference time and energy cost, our work contributes to limited resource computing and sustainable AI technology. © 2020 IEEE.

关键词： Energy utilization

来源：评论

学校读者我要写书评

暂无评论

Innovative Deep Learning Approach for malware Categorization 1

Innovative Deep Learning Approach for malware Categorization

引用

1st International Conference on Computational Intelligence for Security, Communication and Sustainable Development, CISCSD 2024

作者： Lakshmi, B. Lakshmi, B. Maha Pallavi, P. Reddy, M. Pavan Kumar Kumar, B. Naveen Madanapalle Institute of Technology and Science Computer Science and Technology Madanapalle India

ISBN: (纸本)9798350365405

With the improvement of wireless communique, many security threats have emerged. An internet site malware detection gadget (IDS) facilitates come across assaults the business enterprise and the barbarians had been arrested. Earlier, diverse Machine Learning (ML) Methods are implemented to IDS and try to improve the detection outcomes improved accuracy of attackers and IDS. They advocated this article an approach to developing an effective IDS using most important aspect analysis (PCA) and random wooded area category set of rules. Where PCA can assist Organize the facts set by using decreasing the dimensionality of the statistics and make it random a wooded area enables in type. The results received indicate the scope of the technique it plays greater efficiently in terms of accuracy than different such methods SVM, Naive Base and Decision Tree. The result received via the proposed technique Execution time values (min) are 3.24 mins, to be actual (%) is 96.78% and mistakes fee (%) is 0.21%. © 2024 IEEE.

关键词： malware categorization malware detection pca random forest

来源：评论

学校读者我要写书评

暂无评论

Spatio-temporal patterns assisted deep learning model for PM2.5 prediction (STEEP)

引用

Intelligent Data Analysis 2025年第1期29卷 141-155页

作者： Sharmiladevi, S. Siva Sathya, S. Vellore Institute of Technology Chennai India Department of Computer Science Pondicherry University Pondicherry India

Air pollution is an alarming problem in many cities and countries around the globe. The ability to forecast air pollutant levels plays a crucial role in implementing necessary prevention measures to curb its effects in advance. There are many statistical, machine learning, and deep learning models available to predict air pollutant values, but only a limited number of models take into account the spatio-temporal factors that influence pollution. In this study a novel Deep Learning model that is augmented with Spatio-Temporal Co-Occurrence Patterns (STEEP) is proposed. The deep learning model uses the Closed Spatio-Temporal Co-Occurrence Pattern mining (C-STCOP) algorithm to extract non-redundant/closed patterns and the Diffusion Convolution Recurrent Neural Network (DCRNN) for time series prediction. By constructing a graph based on the co-occurrence patterns obtained from C-STCOP, the proposed model effectively addresses the spatiotemporal association among monitoring stations. Furthermore, the sequence-to-sequence encoder-decoder architecture captures the temporal dependencies within the time series data. The STEEP model is evaluated using the Delhi air pollutants dataset and shows an average improvement of 8%–13% in RMSE, MAE and MAPE metric compared to the baseline models. © 2024 – IOS Press. All rights reserved.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Composer Classification Using Maximum Probability Partitioning Based on Compression Principles

引用

Transactions of the Japanese Society for Artificial Intelligence 2024年第2期39卷 1-10页

作者： Takamoto, Ayaka Hironaka, Shiori Umemura, Kyoji Department of Computer Science and Engineering Toyohashi University of Technology Japan Academic Center for Computing and Media Studies Kyoto University Japan

Music classification is a fundamental task in the field of Music Information Retrieval. This paper focuses on composer classification, a specific task within music classification. Compressive techniques are commonly employed in such music classification tasks. In this study, we propose a method to apply the computing Information Quantity using Maximum Probability partitioning to music classification. To evaluate the effectiveness of our proposed method, we perform composer classification, specifically distinguishing between Haydn and Mozart, who are well-known for their stylistic similarities. The experimental results demonstrate that our proposed approach outperforms traditional compression-based classification methods. Furthermore, we compare our method with non-compressive techniques, discussing the significance of feature extraction methods. Our proposed method is a parameter-free classification approach that does not require domain-specific musical expertise or feature extraction based on such expertise. © 2024, Japanese Society for Artificial Intelligence. All rights reserved.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Silver Lining in the Fake News Cloud: Can Large Language Models Help Detect Misinformation?

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年第1期6卷 14-24页

作者： Kumar, Raghvendra Goddu, Bhargav Saha, Sriparna Jatowt, Adam Indian Institute of Technology Department of Computer Science and Engineering Patna801106 India University of Innsbruck Department of Computer Science Innsbruck6020 Austria

In the times of advanced generative artificial intelligence, distinguishing truth from fallacy and deception has become a critical societal challenge. This research attempts to analyze the capabilities of large language models (LLMs) for detecting misinformation. Our study employs a versatile approach, covering multiple LLMs with few- and zero-shot prompting. These models are rigorously evaluated across various fake news and rumor detection datasets. Introducing a novel dimension, we additionally incorporate sentiment and emotion annotations to understand the emotional influence on misinformation detection using LLMs. Moreover, to extend our inquiry, we employ ChatGPT to intentionally distort authentic news as well as human-written fake news, utilizing zero-shot and iterative prompts. This deliberate corruption allows for a detailed examination of various parameters such as abstractness, concreteness, and named entity density, providing insights into differentiating between unaltered news, human-written fake news, and its LLM-corrupted counterpart. Our findings aspire to furnish a refined framework for discerning authentic news, human-generated misinformation, and LLM-induced distortions. This multifaceted approach, utilizing various prompt techniques, contributes to a comprehensive understanding of the subtle variations shaping misinformation sources. © 2024 IEEE.

关键词： Iterative methods

来源：评论

学校读者我要写书评

暂无评论

Privacy and Security Concerns in Edge-Cloud Systems: An In-Depth Analysis 3

Privacy and Security Concerns in Edge-Cloud Systems: An In-D...

引用

3rd International Conference on Intelligent Data Communication Technologies and Internet of Things, IDCIoT 2025

作者： Kaki, Samuel Suhas Kantareddygari, Jagadeswara Reddy Aishwarya, R. School of Computing Sathyabama Institute of Science and Technology Chennai India

ISBN: (纸本)9798331527549

Edge-Cloud systems provide efficient computation and storage close to data sources, with lower latency, scalability, and application performance for various applications, such as IoT, autonomous vehicles, and real-time analytics. These systems are highly vulnerable to significant security and privacy challenges, which include the following: data confidentiality and integrity in the process of transmission, secure protection of sensitive data stored on the edge against physical and remote attacks, and maintaining availability under dynamic network conditions. More challenges related to ensuring users' private nature through anonymizing data, obtaining consent for all processes involved, and compliance with specific regulations like GDPR and CCPA. This work incorporates more enhanced cryptographic methods in the transfer of data between heterogeneous systems and environment-based anomaly detection, advanced privacy-preserving analytics to avoid extraction of private data, adaptive control mechanisms in place for Edge-Cloud for minimizing vulnerability with enhanced trust along with the maintenance of data integrity. The key benefits would be robust mitigation of threats, better interoperability between edge and cloud platforms, and alignment with evolving security standards, which promotes wider adoption across various sectors. © 2025 IEEE.

关键词： Cloud platforms

来源：评论

学校读者我要写书评

暂无评论

An Insightful Approach to Cardiomegaly through CNN and Feature Mapping 1

An Insightful Approach to Cardiomegaly through CNN and Featu...

引用

1st International Conference on Computational Intelligence for Security, Communication and Sustainable Development, CISCSD 2024

作者： Manikandan, R. Mythili, B. Pushyaraga, R. Reddy, M. Karthik Reddy, B. Mallikarjun Madanapalle Institute of Technology and Science Computer Science and Technology Madanapalle India

ISBN: (纸本)9798350365405

The study presents a new approach to identifying cardiomegaly using advanced convolutional neural networks, more specifically, MobileNet and VGG architectures, with interpretative feature maps. Cardiomegaly, or an enlarged heart, is an important sign of particular cardiovascular disorders. Consistently, this topic is an early finding in individual x-ray projection, and because of the general precision of classical methods, innovative deep learning algorithms are needed. In the first stage, the MobileNet and VGG models are designed on distinct heart photo galleries used to cardiac image classification. Systems are seen throughout the domain as vital utilities designed for photo classification. Ultimately, to simplify the solution, an interpreted feature map is employed. Therefore, the technique becomes more reliable, and the design presents a quantitative statement of whether the accessible results say anything concerning the causes of the cardiomegaly estimates. This research has been extensively tested with datasets and produces better results than the existing methods. The presented model grammar map is interpretable, which aids in the medical professional's comprehension of the model's prioritizes and boosts their self-assurance in the diagnostic procedure. Furthermore, the novel approach is versatile with various imaging modalities and patient populace, which suggests extensive practical plays. With the help of an interpretable feature map and the combined usage of VGG appearances and MobileNet advantages, this work complements to the field of substantiation the identification of cardiomegaly. As a final point, the introduced model acquires an excellent amount of truthfulness and resolves a serious predicament of interpretability. Thus, it satisfies the growing stress on lucidness and dependability of deep learning applications in medical substantiation. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：